Where do you start your mixes from?

Nola

Nola

Well-known member
Hey guys, just curious to hear your mix process. Especially those of you who are pros doing this for a living. Where do you start the mix and why? do you start with the drums/bass, the vocals, etc?

Also, say you start with the drum or vocal -- how do you decide what volume to start them at?

I have my own process and it goes okay, but I feel like it takes too long to mix a song because I kind of jump around too much and never know where to start and at what volume. I know the idea is to highly the key points and make everything else support it. I understand those concepts but still can't get a quicker work flow.

Last night was an awakening moment b/c after finishing a mix I realized I had to re-mix the kick and bass a little differently and when I did all my automation no longer sounded good, so I might have to redo everything. This can't be the right way.

Thanks
 
Last edited:
I start with the drums. I set up a drum bus and a overhead bus and route everything appropriately.

I do rock and metal, so the drums need to be agressive, so I compress the overheads and the drum buss.

I record the drums so that they tend to peak around -6dbfs. I start with all the channel faders at unity and insert the channel EQ's and compressors and start working on the individual drum sounds. I will normally do any EQing on the overheads at the bus, so I don't have to match two EQ's on the individual channels.

Once I am done dialing the drums in, they are what they are and I just fill everything in around them. A normal 30 track mix will end up peaking around -6dbfs to -3dbfs. But it absolutely doesn't matter, as long as it doesn't clip.

The final volume of the song doesn't really happen at the mix stage.
 
Farview said:
I start with the drums. I set up a drum bus and a overhead bus and route everything appropriately.

I do rock and metal, so the drums need to be agressive, so I compress the overheads and the drum buss.

I record the drums so that they tend to peak around -6dbfs. I start with all the channel faders at unity and insert the channel EQ's and compressors and start working on the individual drum sounds. I will normally do any EQing on the overheads at the bus, so I don't have to match two EQ's on the individual channels.

Once I am done dialing the drums in, they are what they are and I just fill everything in around them. A normal 30 track mix will end up peaking around -6dbfs to -3dbfs. But it absolutely doesn't matter, as long as it doesn't clip.

The final volume of the song doesn't really happen at the mix stage.
Click to expand...

Thanks, Fairview, that's exactly what I'm looking for. -6db gives me an idea at least on drums. I like to start with them, but never sure on the volume I should be shooting for, and it's been messing up mixes.

When you put compression on say a vocal it might get louder or more upfront, so when you start mixing in vocals or instruments that use compression, do you aim for a certain reduction (like "I want -5db compression on bass") and set that or do you compress bass and vocal by ear in relation to drums? Do you bring in bass and vocal after drums? Then guitars last?

So basically in a rock mix having everything in relation to the drums is what you recommend? That's how I do it now. Last night I decided the low end was off and took out some kick and let the bass carry the low end more. When I did that, the entire mix changed and all the automation I had done was off on the guitars now. Is this just the nature of the beast?
 
So I just opened a mix and looked at the drums, and they're peaking at -17db. So that's way lower than yours. That's with the fader at unity (and they're samples).
So should I push the fader up above unity to get it closer to -6db peaks?

I feel like if I do that all my faders will be pushed up above unity just trying to get them to sit above the drums.
 
I usually start with the kick, setting the fader to 0 then using clip gain to get it to peak around -16dBFS. Then snare, fader at 0, mixing it to the kick with clip gain. Since snare is generally in the overheads I include them as part of mixing the snare to the kick. Then lead vocal, then bass, then everything else.
 
I usually start with the drums, this is only so I can get the drums balanced enough to put in the mix, then I add bass. From there I get the rest of the rhythm section together and then rough mix the band with no vocals.

May sound strange to some as the vocal is the most important. However I then zero in the vocal, placing it in the mix, seeing how it interacts, soloing the vocal playing with the EQ, compression, etc and putting it back into the band mix. When I have the vocals really working I then go back to the band mix / overall mix, fine tune, and check that nothing in the band is upsetting the vocals. Then it goes around and around until I am happy with it, eq, effects, volume changes throughout the song etc etc.

Alan.
 
Nola said:
Thanks, Fairview, that's exactly what I'm looking for. -6db gives me an idea at least on drums. I like to start with them, but never sure on the volume I should be shooting for, and it's been messing up mixes.

When you put compression on say a vocal it might get louder or more upfront, so when you start mixing in vocals or instruments that use compression, do you aim for a certain reduction (like "I want -5db compression on bass") and set that or do you compress bass and vocal by ear in relation to drums? Do you bring in bass and vocal after drums? Then guitars last?
Click to expand...
I compress for the sound of compression, not necessarily to tame the dynamics. I compress something until it sounds right, then I place it in the mix. It sounds like you may be doing the opposite.

I get the bass sounding right and put it with the drums. Once they are rocking together, I bring up the guitars. Then I add vocals. But again, I compress to get a sound, so once I get a sound I just place it where it needs to be. If something is clashing, I tend to fix it with EQ.

Nola said:
So basically in a rock mix having everything in relation to the drums is what you recommend? That's how I do it now. Last night I decided the low end was off and took out some kick and let the bass carry the low end more. When I did that, the entire mix changed and all the automation I had done was off on the guitars now. Is this just the nature of the beast?
Click to expand...
Kind of. What I do is get a mix that is really solid before I do any automation. That will keep you from chasing your tail. I also try to give everything enough room to move, which keeps me from having to automate so much.

For example, if you need to make the bass more apparent, add some 800hz instead of turning it up. The low end will stay where it is, but the bass will seem louder in the mix. Shaping the sounds with EQ can give everything it's own space in the spectrum, so it can all be loud, but not fight.
 
Nola said:
So I just opened a mix and looked at the drums, and they're peaking at -17db. So that's way lower than yours. That's with the fader at unity (and they're samples).
So should I push the fader up above unity to get it closer to -6db peaks?

I feel like if I do that all my faders will be pushed up above unity just trying to get them to sit above the drums.
Click to expand...
I wouldn't do it at the fader, I would turn up the actual clips. But, if everything you record has these sorts of levels, there isn't any real reason why you need to do that.

Don't worry about fader position. Where the faders end up has to do with how hot you recorded the tracks vs. where they need to sit in the mix relative to the other tracks.

In the future, you maybe should record the drums a little hotter. If they are midi instruments, you might want to adjust the velocity. Hitting them harder might make the drums sound better, depending on the samples you are using.
 
I'll have to say the same thing...it's always based on the drums.
I don't much pay attention to where the faders are...but most of that is taken care of during tracking. I make sure I get the right levels down...that way, if I put the faders at unity when mixing, the drums are going to be sounding pretty much how they should, and as loud as they should. I might tweak the Kick/Snare...and the HH...but they're generally where they need to be.

After that...the rest of the tracks are brought up relative to the drums...usually the backing tracks next...and then finally the lead tracks.
 
I try to start with everything except the "sweetening" (reverb/FX, solos, BG Vox, featured oddball instruments, etc.) up at nominal levels. I'll listen to the whole track getting a general impression of how well it works as a song/arrangement. Are 2 instruments fighting for space? Are there enough dynamic changes coming from the performance/arrangement, or will I have to address that in the mix? Is there something missing? Something that's superfluous? Once I have a feel for that, I'll start panning, then approach the E.Q. (I never EQ in "solo" only in the mix, though I WILL listen in solo - usually to 2 parts at once that conflict - as that gives me a better idea what to adjust.) Then I'll add the sweetening parts one at a time, and decide if any adjustments to the "static" mix will be needed, or if I will just adjust the new parts.
 
Like most, I start with drums, then bring up bass. Once they are sitting well together I then go for the lead vocals.

What I aim for is a mix of drums bass and vocals that could pass as a final mix . . . they should all work nicely together. Only when I get this do I add in other instruments, then other vocals.

I usually mix with effects (reverb etc), but this can change as the mix progresses. I always mix in stereo.
 
Hi Nola. Fairview is giving some pretty good advice here.

When I build from the ground up (as in starting with bass and drums then adding from there), I try to get my tone shaping and surgical plugins working before really messing with the level. In other words, I'm trying to get the whole thing under control. Flabby shitting sounding kick drums, boxy, or potato chip sounding snares, overly abrasive cymbals. Start by getting it under control. I start with the kick and move immediately to overheads, room, and ambience mics. That's 80% of your sound. Then use your snare and tom mics to fill in the missing gaps. And you are not obligated to use them. If your overheads are doing all the work, and doing it sufficiently, then just leave the toms mics out. Get your gates and samples set on the drums asap. You'll know right away weather you want the bleed in, or want it gated out. I try to match the bass against the kick, but not until I have it processed. I don't need the meters quite as much because my room is accurate and I have one of the best monitoring chains money can buy. But if you don't, I'd probably recommend soloing your kick to where its hitting -3dbvu. Then solo the bass and adjust the level to where it hits at 0dbvu when combined with the kick. That'll get you close. Use your ear from there.
 
Whether or not your overheads account for 80% of your drum sound really depends on the style of music you are doing. Since I do a lot of metal, the room mic sound just won't work. For me, when I'm doing the larger than life over the top metal thing, the only thing I'm getting from overheads and room mics are cymbals, reverb and part of the snare sound.

I also want to point out that jkuehlin was giving db values in dbVU, not dbFS. Unfortunately, when dealing with drums, there is no good way to translate between the two scales. VU meters have a slow response time, so they are giving you more or less an average level. Meters that measure dbFS are peak meters and are very fast, so they catch all the transients that VU meters will miss. Since drums are mostly transients, the same drum hit can measure less than 0dbVU while still clipping above 0dbFS. Instruments with more sustain, like bass, will have a sustained note at 0dbVU that will be -18dbFS in the computer.
 
Drums
-kick
-snare
-toms and misc kit percussion
-OHs
-other percussion
bass
lead vocals
rhythm guitars
lead guitars
other instruments
bgvox

The idea being that the drums, bass, and lead vox are the "core". The rhythm guitars provide the body and tone. Everything else is the fluff that makes it interesting.
 
I usually start with getting a rough mix of the drums for panning and levels. I try to get them to a good starting point to build on. Then I add guitars (electric or acoustic), then bass guitar. Lead vocals is next, then lastly I add backing vocals. I get a rough mix of the song going with each instrument added, then I go back and make tweaks where needed to get the polished sound I'm looking for. This is what I've found to be the easiest for me.
 
Farview said:
Whether or not your overheads account for 80% of your drum sound really depends on the style of music you are doing. Since I do a lot of metal, the room mic sound just won't work. For me, when I'm doing the larger than life over the top metal thing, the only thing I'm getting from overheads and room mics are cymbals, reverb and part of the snare sound.

I also want to point out that jkuehlin was giving db values in dbVU, not dbFS. Unfortunately, when dealing with drums, there is no good way to translate between the two scales. VU meters have a slow response time, so they are giving you more or less an average level. Meters that measure dbFS are peak meters and are very fast, so they catch all the transients that VU meters will miss. Since drums are mostly transients, the same drum hit can measure less than 0dbVU while still clipping above 0dbFS. Instruments with more sustain, like bass, will have a sustained note at 0dbVU that will be -18dbFS in the computer.
Click to expand...

Hmmm. Interesting. I wouldn't know. I've mixed some heavy rock stuff but never strait up metal. I'd turn the job down and send it to someone who knows what they're doing.
 
I wasn't trying to call you out or anything, I was just trying to point out that different genres require different approaches to the instruments in the mix. The drums in a jazz mix are expected to do a different job than the drums on a Metallica album...same with most of the other instruments as well.

The 'room and overheads being most of the sound' thing works in any form of music where the drums are expected to sound like the drums actually sound in the room. When they need to be huge, in your face, and not at all what a drum would really sound like standing next to it, close mics and a ton of EQ and compression are the way to go.
 
^^^See that's where you start to lose me. If they're not supposed to sound like real drums, then why are you starting with real drums? Especially if the "performance" is going to get edited onto a grid, smashed to hell for dynamic consistency, and layered with samples anyway. MIDI makes timing and dynamic edits super easy, and if you just pull up samples that sound the way you want to begin with...
 
I guess because I was doing it before computers... most of the editing to the grid and sample replacement is a function of low budgets/time constraints.

The example of the sort of understanding I'm talking about would be like the metallica black album. That uses the rooms as verb, but relies on the close mics and album shit ton of eq to get it to sound like that.
 
Farview said:
I wasn't trying to call you out or anything, I was just trying to point out that different genres require different approaches to the instruments in the mix. The drums in a jazz mix are expected to do a different job than the drums on a Metallica album...same with most of the other instruments as well.
Click to expand...
all good...I didn't take it like that. And I'll be the first to admit I have very little experience with that style. I am surprised you referred to Metallica as metal...I tend to think of them as radio pop/top 40. Or classic rock maybe. I thought you were talking about sludge death screamo tear-your-head-off ish stuff. I figured that in a mix with all that noise, room mics might be pretty lost on the overall sound.
The 'room and overheads being most of the sound' thing works in any form of music where the drums are expected to sound like the drums actually sound in the room. When they need to be huge, in your face, and not at all what a drum would really sound like standing next to it, close mics and a ton of EQ and compression are the way to go.
Click to expand...

Foo Fighters, Nickelback, Korn, Creed...room mics doing all the work. I have tracked sessions with Josh Freese and Kenny Arnoff on drums. The close mics are hardly doing shit. I posted isolated samples of the drums as well as my signal chains on the thread about the drum busses. That was one of the sessions with Kenny.

https://homerecording.com/bbs/general-discussions/mixing-techniques/setting-up-drum-bus-391866/

there's the link. **The mp3 of the drums by themselves is in post #2 on that thread below the pictures.
 
You must log in or register to reply here.
Back
Top